make sparsevec preserve zeros

[KristofferC]

After #15242 there is a mismatch between the constructors of sparse matrices / sparse vectors whether they preserve values in the sparse structure that happened to accumulate to zero.

Currently, matrices keeps them while vectors remove them:

julia> sparsevec([1,2,3], [0.0, 1.0, 2.0])
Sparse vector of length 3, with 2 Float64 nonzero entries:
  [2]  =  1.0
  [3]  =  2.0


julia> sparse([1, 2, 3], [1, 1, 1], [0.0, 1.0, 2.0])
3x1 sparse matrix with 3 Float64 entries:
    [1, 1]  =  0.0
    [2, 1]  =  1.0
    [3, 1]  =  2.0

This makes sparse vectors behave the same as matrices:

julia> sparsevec([1,2,3], [0.0, 1.0, 2.0])
Sparse vector of length 3, with 3 Float64 nonzero entries:
  [1]  =  0.0
  [2]  =  1.0
  [3]  =  2.0

[KristofferC]

Bump.

[KristofferC]

bump